I think we'd need a "healthy years lost" analysis to see this. Take the residual expected lives of each cohort and guess how many of those years are lost due to

I think we'd need a "healthy years lost" analysis to see this. Take the residual expected lives of each cohort and guess how many of those years are lost due to the illness. Similarly some kind of measure for the weeks in hospital and potential permanent damage from surviving.